The present invention relates to the field of electronics and power operational amplifiers and, more specifically, to the variety of power operational amplifiers that are described as high voltage power operational amplifiers.
In Prior Art FIGS. 1 and 2 and also in FIGS. 3 and 4, which illustrate the invention, multiple amplifier stages that represent the second stage and the output stage of an operational amplifier are illustrated. Node 2 is a signal from the differential input stage of the operational amplifier. The input stage design is well understood by those skilled in the art and is not required for understanding of the present invention, and is so omitted. The second stage is labeled as such because it is very common in the design of operational amplifiers to have three stages: a differential input stage, a second stage and an output stage. Each of the figures also shows several MOSFET transistors stacked vertically drain-to-source between the positive supply voltage and the negative supply voltage. Although a total of four MOSFETs are shown stacked in each stage, the description herein could equally be applied to stacks of any number. Also, although the figures and explanations apply to MOSFET transistor amplifier designs, the techniques are not limited to this type of technology. Similar circuits could be demonstrated with bipolar transistors, for example.
In Prior Art FIG. 1, four resistors, R1, R2, R3, and R4 are connected in series across the power supply voltages. Although all the resistors need not be equal in value, they will be assumed to be for explanation purposes. For the illustration of circuit operation the positive and negative supply voltages are assumed to be +400V and xe2x88x92400V respectively. However, the supply voltages could be any desired value, as the circuit would still operate in a similar fashion. Also, it is noted that Q7 in the figures is a type of transistor known as a depletion mode MOSFET and that it is connected as a current source. The exact nature of the current source is not important for circuit operation, but there needs to be a current source at the Q7 location. Typically a transistor of similar type to the others is used in most designs. R1, R2, R3 and R4 along with capacitors C1, C2, C3 and C4 (to be discussed later) will hereinafter be referred to as xe2x80x9cthe biasing stringxe2x80x9d.
Assuming equal resistor values, each resistor will have xc2xc of the total supply voltage dropped across it. Since little or no current flows into the gate terminal of a MOSFET transistor +200V will appear on the gate of Q3, zero volts at circuit node 6, and xe2x88x92200V will appear on the gate of Q5 (all with respect to ground, of course). These biasing voltages force the source terminal of Q3 to be approximately +200V as well since only a few volts of bias voltage from gate to source is required to operate a MOSFET. Similarly the source of Q5 is required to be about +200V. It is evident that the bias applied by the four resistors is approximately reflected in the drain-source voltages of Q1, Q3, Q5 and Q7.
Note that in an operational amplifier some feedback is applied from the output to the differential stage input. If the differential stage input signal is set to zero volts then the output voltage will be at zero volts as well (neglecting offset errors). For simplicity all of the figures do not show the output connected to a feedback network. The terminal labeled xe2x80x9cinputxe2x80x9d is a signal that comes from the previously mentioned differential input stage that would be required to complete an operational amplifier design. In each figure, the schematic should be assumed to be merely a portion of that overall design.
Referring again to Prior Art FIG. 1, if the differential input stage (not shown) were to command the output (via the input terminal) to move to +50V then node 6 would have to move to +50V plus whatever gate-source voltage would be required to operate Q4. As the output moves, node 6 is pushed to a corresponding voltage. Suppose, for example, the output moves to +350V. Node 6 will follow to approximately the same voltage. R3, R4 divides that voltage and as previously explained Q5 and Q7 will now have approximately equal voltages distributed across each from drain to source (175V). Suppose further that the transistors used in this example had a breakdown voltage of only 200V. With this technique a total voltage supply of 400V could be used.
This technique works well in many applications. However, there is an important limitation. So far, the values of resistors R1-R4 have not been considered. In principle, any value could be used. However, the lower the value, the higher the biasing current requirement of the power supplies. This biasing current increases the overall power consumption of the circuit and does not contribute to the output. Higher power supply currents at higher supply voltages are expensive. Consequently, R1-R4 is usually set to a value of 1 meg-ohm, or more.
Therein lies the major limitation of this circuit. As the output moves, charge must move into or out of the gate circuits of Q3 and Q5. If the values of R1-R4 are high, this cannot be done quickly because of the time constants formed by the capacitance seen at the gates of the transistors and the values of the resistors. For example, if a signal at the input terminal quickly drives the gate of Q1 more negative (Q1 more on), Q""s drain-source voltage will quickly decrease. If the gate voltage on Q3 cannot quickly follow that voltage change, the voltage that should be equally divided across Q3 and Q1 will mostly appear across Q3. This unequal distribution of voltage will destroy Q3 since its breakdown voltage will be exceeded. An unequal distribution of voltage across output transistors Q2 and Q4 will also result, with the same damaging effects.
Prior Art FIG. 2 shows a method to improve the speed of operation of the circuit compared to FIG. 1. In FIG. 2 capacitors C1, C2, C3 and C4 are added to the biasing string. The capacitors allow more charge to dynamically move in and out of the gate circuits in less time than would otherwise be possible without them, as discussed above when considering Prior Art FIG. 1. This method is an improvement but not without problems as well. The same capacitors that improve the circuit also build in a limitation. Using small signal AC analysis of node 6, capacitors C1, C2, C3 and C4 appear as an equivalent capacitor from node 6 to ground. The equivalent capacitor rolls off the high frequency response of the second stage, thus limiting the bandwidth of the amplifier. That capacitance is charged by the current source Q7. Current that charges the equivalent capacitor is not then available to charge other capacitance in the stage. Thus, these capacitors also limit the slew rate of the amplifier since current available to charge these capacitances is what sets the slew rate of the amplifier.
Further, as node 6 moves more negative, the current required by R1 and R2 in the divider increases since the voltage drop across them increases. That current comes from current source Q7. That leaves less current to charge the equivalent capacitance on node 6 further limiting the slew rate of the amplifier.
Accordingly, conventional operational amplifiers suffer from unbalanced distribution of voltages, for example, due to increased time constants or limited slew rates. Additionally, conventional amplifier circuits provide inefficient power (current) consumption, since these effects are not taken into design considerations Thus, there exists a need to provide an operational amplifier having more efficient power considerations, as well as, increased slew rate and increased speed.
Accordingly, the present invention provides an improved means to bias a second stage and an output stage where the total supply voltage exceeds the breakdown voltage of the individual transistors used to build the several stages of the amplifier. In Prior Art FIGS. 1 and 2 existing methods are shown to distribute the supply voltages across several transistors which are stacked drain-to-source in a serial arrangement so as to increase the total breakdown voltage of the amplifier. The problems with the Prior Art models relating to the rate at which the amplifier""s signal can change without causing unequal distribution of voltages across the individual transistors will be discussed, and it will be shown how the present invention solves this problem.
In one embodiment the present invention provides an amplifier circuit comprising: a positive supply rail and a negative supply rail; a first plurality of switches stacked between said positive and negative supply rails, a biasing string comprising a plurality of RC circuits stacked between said positive and negative supply rails, one of said switches receiving an external control signal to control the conduction state thereof, another of said switches being arranged as a current source; a second plurality of switches stacked between said positive and negative supply rails the conduction state thereof being controlled by the conduction state of said first plurality of switches; and a feedback line from said second plurality of switches to said biasing string operable to drive said plurality of RC circuits using power from said second plurality of switches.
In a second embodiment the present invention an amplifier circuit comprising a positive supply rail and a negative supply rail; a first plurality of switches stacked between said positive and negative supply rails, a biasing string comprising a plurality of RC circuits stacked between said positive and negative supply rails, one of said switches receiving an external control signal to control the conduction state thereof, another of said switches being arranged as a current source; a second plurality of switches stacked between said positive and negative supply rails the conduction state thereof being controlled by the conduction state of said first plurality of switches; and a feedback line from said second plurality of switches to said first plurality of switches to bias said first plurality of switches with said second plurality of switches.
In broader aspects, the present invention provides an amplifier circuit, comprising an input stage operable to produce a differential input signal; a second stage operable to amplify said input signal; an output stage operable to deliver said amplified signal to a load; and a local feedback line between said output stage and said second stage, said feedback line operable to bias said second stage with the potential of said output stage.
The present invention also includes methods to amplify a signal, comprising the steps of: passing an input signal through an amplifier stage comprising a plurality of first switches stacked between positive and negative power supply rails; controlling the conduction state of said first switches to amplify and pass said signal to an output stage comprising a plurality second switches stacked between said positive and negative power supply rails; providing a local feedback line between said amplifier stage and said output stage so that said output stage controls the biasing of said amplifier stage.
Other features and advantages of the present invention will become apparent as the following Detailed Description proceeds, and upon reference to the Drawings, wherein like numerals depict like parts, and wherein: